ams journals online - journal - table of contents · ams journals online - journal - table of...

Click here to load reader

Post on 11-Apr-2018

216 views

Category:

Documents

4 download

Embed Size (px)

TRANSCRIPT

  • 10-12-2 7:43AMS Journals Online - Journal - Table of Contents

    1 2 http://journals.ametsoc.org/toc/clim/23/18

    INST OF ATMOS. PHYSICS Sign In or Institutional Administrator | Help

    Quick Search Full Text Journals Subscribe For Authors Information Online Help

    Next >< Previous

    Volume 23 Issue 18 (September 2010)

    Current Issue Available Issues

    Early Online Releases Author Index

    Journal Information

    Online ISSN: 1520-0442 Print ISSN: 0894-8755 Frequency: Semimonthly

    Special Collections Staff and Editors Instructions to Authors Manuscript Submission How to Subscribe

    AMS Alerts

    (What is this?)JOIN: Journal Online IssueNotification

    Journal of Climate

    Climate research concerned with large-scale variability of the atmosphere, oceans, and land surface, including the cryosphere; past, present andprojected future changes in the climate system (including those caused by human activities); climate simulation and prediction. Occasionally theJournal of Climate will publish review articles on particularly topical areas. Such reviews must be approved by the Chief Editor prior tosubmission.

    Table of Contents

    Volume 23, Issue 18 (September 2010)

    View Abstracts Add to Favorites Email Download to Citation Manager Track Citations

    Select All

    ARTICLES

    4717 Australian Monsoon Variability Driven by a GillMatsuno-Type Response to Central West Pacific WarmingAndra S. Taschetto, Reindert J. Haarsma, Alexander Sen Gupta, Caroline C. Ummenhofer, Khalia J. Hill, Matthew H.EnglandAbstract . Full Text . PDF (6695 KB)

    47174736

    4737 Isotopic and Elemental Changes in Winter Snow Accumulation on Glaciers in the Southern Alps of New ZealandHeather Purdie, Nancy Bertler, Andrew Mackintosh, Joel Baker, Rachael RhodesAbstract . Full Text . PDF (1604 KB)

    47374749

    4750 Spatial Structure, Forecast Errors, and Predictability of the South Asian Monsoon in CFS Monthly RetrospectiveForecastsHae-Kyung Lee Drbohlav, V. KrishnamurthyAbstract . Full Text . PDF (4146 KB)

    47504769

    4770 The MaddenJulian Oscillation Simulated in the NCEP Climate Forecast System Model: The Importance ofStratiform HeatingKyong-Hwan Seo, Wanqiu WangAbstract . Full Text . PDF (4109 KB)

    47704793

    4794 Model Fidelity versus Skill in Seasonal ForecastingTimothy DelSole, Jagadish ShuklaAbstract . Full Text . PDF (996 KB)

    47944806

    4807 Asymmetry of Atmospheric Circulation Anomalies over the Western North Pacific between El Nio and La NiaBo Wu, Tim Li, Tianjun ZhouAbstract . Full Text . PDF (2022 KB)

    48074822

    4823 Tracking and Mean Structure of Easterly Waves over the Intra-Americas SeaYolande L. Serra, George N. Kiladis, Kevin I. HodgesAbstract . Full Text . PDF (3033 KB)

    48234840

    4841 Milankovitch Forcing and Meridional Moisture Flux in the Atmosphere: Insight from a Zonally Averaged OceanAtmosphere ModelAndrs Antico, Olivier Marchal, Lawrence A. Mysak, Franoise VimeuxAbstract . Full Text . PDF (1463 KB)

    48414855

    4856 A Pseudoproxy Evaluation of the CCA and RegEM Methods for Reconstructing Climate Fields of the LastMillenniumJason E. Smerdon, Alexey Kaplan, Diana Chang, Michael N. EvansAbstract . Full Text . PDF (21677 KB)

    48564880

    4881 Changes in Precipitation Extremes in the Hawaiian Islands in a Warming ClimatePao-Shin Chu, Ying Ruan Chen, Thomas A. SchroederAbstract . Full Text . PDF (2489 KB)

    48814900

    4901 The NCEP GODAS Ocean Analysis of the Tropical Pacific Mixed Layer Heat Budget on Seasonal to InterannualTime ScalesBoyin Huang, Yan Xue, Dongxiao Zhang, Arun Kumar, Michael J. McPhadenAbstract . Full Text . PDF (2278 KB)

    49014925

    4926 Exploring AtmosphereOcean Coupling Using Principal Component and Redundancy AnalysisFaez Bakalian, Harold Ritchie, Keith Thompson, William MerryfieldAbstract . Full Text . PDF (3450 KB)

    49264943

    4944 Asymmetry in ENSO Teleconnection with Regional Rainfall, Its Multidecadal Variability, and ImpactWenju Cai, Peter van Rensch, Tim Cowan, Arnold SullivanAbstract . Full Text . PDF (4287 KB)

    49444955

    4956 TimeFrequency Characteristics of Regional Climate over Northeast China and Their Relationships with

    All Journals > Journal of Climate > September 2010 Advanced Search

    http://journals.ametsoc.org/action/ssostart?idp=https%3A%2F%2Fshib.ametsoc.org%2Fshibboleth%2Fidphttps://journals.ametsoc.org/action/showLogin?uri=%2Ftoc%2Fclim%2F23%2F18http://journals.ametsoc.org/helphttp://journals.ametsoc.org/http://www.ametsoc.org/pubs/subscribe/index.htmlhttp://www.ametsoc.org/pubs/arcindex.htmlhttp://journals.ametsoc.org/page/informationhttp://journals.ametsoc.org/helphttp://journals.ametsoc.org/toc/clim/23/19http://journals.ametsoc.org/toc/clim/23/17http://journals.ametsoc.org/toc/clim/23/21http://journals.ametsoc.org/loi/climhttp://journals.ametsoc.org/toc/clim/0/0http://journals.ametsoc.org/action/showAuthorIndex?journalCode=climhttp://journals.ametsoc.org/page/collectionhttp://www.ametsoc.org/EXEC/contacts.html#DIRPhttp://www.ametsoc.org/pubs/arcindex.htmlhttp://www.ametsoc.org/au_upload/index.cfmhttp://www.ametsoc.org/pubs/subscribe/index.htmlhttp://journals.ametsoc.org/page/feeds.jsphttp://journals.ametsoc.org/help?context=rsshttp://www.ametsoc.org/pubs/listserve/getlistserve_journal.cfmjavascript:submitArticles(document.frmAbs,%20'/action/showMultipleAbstracts')javascript:submitArticles(document.frmAbs,%20'/action/showMailPage')javascript:submitArticles(document.frmAbs,%20'/action/showCitFormats')http://journals.ametsoc.org/doi/abs/10.1175/2010JCLI3474.1http://journals.ametsoc.org/doi/full/10.1175/2010JCLI3474.1http://journals.ametsoc.org/doi/pdf/10.1175/2010JCLI3474.1http://journals.ametsoc.org/doi/abs/10.1175/2010JCLI3701.1http://journals.ametsoc.org/doi/full/10.1175/2010JCLI3701.1http://journals.ametsoc.org/doi/pdf/10.1175/2010JCLI3701.1http://journals.ametsoc.org/doi/abs/10.1175/2010JCLI2356.1http://journals.ametsoc.org/doi/full/10.1175/2010JCLI2356.1http://journals.ametsoc.org/doi/pdf/10.1175/2010JCLI2356.1http://journals.ametsoc.org/doi/abs/10.1175/2010JCLI2983.1http://journals.ametsoc.org/doi/full/10.1175/2010JCLI2983.1http://journals.ametsoc.org/doi/pdf/10.1175/2010JCLI2983.1http://journals.ametsoc.org/doi/abs/10.1175/2010JCLI3164.1http://journals.ametsoc.org/doi/full/10.1175/2010JCLI3164.1http://journals.ametsoc.org/doi/pdf/10.1175/2010JCLI3164.1http://journals.ametsoc.org/doi/abs/10.1175/2010JCLI3222.1http://journals.ametsoc.org/doi/full/10.1175/2010JCLI3222.1http://journals.ametsoc.org/doi/pdf/10.1175/2010JCLI3222.1http://journals.ametsoc.org/doi/abs/10.1175/2010JCLI3223.1http://journals.ametsoc.org/doi/full/10.1175/2010JCLI3223.1http://journals.ametsoc.org/doi/pdf/10.1175/2010JCLI3223.1http://journals.ametsoc.org/doi/abs/10.1175/2010JCLI3273.1http://journals.ametsoc.org/doi/full/10.1175/2010JCLI3273.1http://journals.ametsoc.org/doi/pdf/10.1175/2010JCLI3273.1http://journals.ametsoc.org/doi/abs/10.1175/2010JCLI3328.1http://journals.ametsoc.org/doi/full/10.1175/2010JCLI3328.1http://journals.ametsoc.org/doi/pdf/10.1175/2010JCLI3328.1http://journals.ametsoc.org/doi/abs/10.1175/2010JCLI3484.1http://journals.ametsoc.org/doi/full/10.1175/2010JCLI3484.1http://journals.ametsoc.org/doi/pdf/10.1175/2010JCLI3484.1http://journals.ametsoc.org/doi/abs/10.1175/2010JCLI3373.1http://journals.ametsoc.org/doi/full/10.1175/2010JCLI3373.1http://journals.ametsoc.org/doi/pdf/10.1175/2010JCLI3373.1http://journals.ametsoc.org/doi/abs/10.1175/2010JCLI3388.1http://journals.ametsoc.org/doi/full/10.1175/2010JCLI3388.1http://journals.ametsoc.org/doi/pdf/10.1175/2010JCLI3388.1http://journals.ametsoc.org/doi/abs/10.1175/2010JCLI3501.1http://journals.ametsoc.org/doi/full/10.1175/2010JCLI3501.1http://journals.ametsoc.org/doi/pdf/10.1175/2010JCLI3501.1http://journals.ametsoc.org/http://journals.ametsoc.org/loi/climhttp://journals.ametsoc.org/search/advancedwubo

  • Asymmetry of Atmospheric Circulation Anomalies over the WesternNorth Pacific between El Nino and La Nina*

    BO WU

    LASG, Institute of Atmospheric Physics, Chinese Academy of Sciences, and Graduate University

    of Chinese Academy of Sciences, Beijing, China

    TIM LI

    IPRC, and Department of Meteorology, University of Hawaii at Manoa, Honolulu, Hawaii

    TIANJUN ZHOU

    LASG, Institute of Atmospheric Physics, Chinese Academy of Sciences, Beijing, China

    (Manuscript received 27 April 2009, in final form 26 January 2010)

    ABSTRACT

    The asymmetry of the western North Pacific (WNP) low-level atmospheric circulation anomalies between

    the El Nino and La Nina mature winter is examined. An anomalous WNP cyclone (WNPC) center during La

    Nina tends to shift westward relative to an anomalous WNP anticyclone (WNPAC) center during El Nino.

    Two factors may contribute to this asymmetric response. The first factor is the longitudinal shifting of El Nino

    and La Nina anomalous heating. The composite negative precipitation anomaly center during La Nina is

    located farther to the west of the composite positive precipitation anomaly center during El Nino. The

    westward shift of the heating may further push the WNPC westward relative to the position of the WNPAC.

    The second factor is the amplitude asymmetry of sea surface temperature anomalies (SSTAs) in the WNP,

    namely, the amplitude of local cold SSTA during El Nino is greater than that of warm SSTA during La Nina.

    The asymmetry of SSTA is originated from the asymmetric SSTA tendencies during the ENSO developing

    summer. Although both precipitation and surface wind anomalies are approximately symmetric, the surface

    latent heat flux anomalies are highly asymmetric over the key WNP region, where the climate mean zonal

    wind speed is small. Both the anomalous westerly during El Nino and the anomalous easterly during La Nina

    in the region lead to an enhanced surface evaporation, strengthening (weakening) the enhancement of the

    cold (warm) SSTA in situ during El Nino (La Nina). The asymmetry of the SSTA in the WNP is further

    amplified due to anomalous wind differences between El Nino and La Nina in their mature winter. Atmo-

    spheric general circulation model experiments demonstrate that both factors contribute to the asymmetry

    between the WNPAC and WNPC. The asymmetric circulation in the WNP contributes to the asymmetry of

    temporal evolutions between El Nino and La Nina.

    1. Introduction

    During the mature phase of El Nino, the most prom-

    inent low-level atmospheric circulation anomalies over

    the tropical western Pacific and East Asia is an off-

    equatorial anomalous anticyclone [hereafter the western

    North Pacific (WNP) anticyclone (WNPAC)] (Wang et al.

    2000; Chang et al. 2000a). The WNPAC is a key bridge

    that links El Nino and the East Asian summer monsoon

    (Chang et al. 2000a,b; Wang et al. 2000; Wu et al. 2003).

    It influences the East Asian climate from the El Nino

    mature winter to the following summer. Concurring with

    the formation of the WNPAC, a positive anomalous rain-

    fall belt extends from southeastern China to the Kuroshio

    Extension region in boreal winter and spring owing to

    anomalous moisture transport by the southerly compo-

    nent in the northwestern flank of the WNPAC (Zhang

    * School of Ocean and Earth Science and Technology Contri-

    bution Number 7907 and International Pacific Research Center

    Contribution Number 683.

    Corresponding author address: Dr. Tianjun Zhou, LASG, In-

    stitute of Atmospheric Physics, Chinese Academy of Sciences,

    Beijing 100029, China.

    E-mail: [email protected]

    15 SEPTEMBER 2010 W U E T A L . 4807

    DOI: 10.1175/2010JCLI3222.1

    2010 American Meteorological Society

  • et al. 1996; Lau and Nath 2000; Zhang and Sumi 2002).

    During the subsequent summer, the WNPAC favors a

    westward extension of the western Pacific subtropical

    high, which blocks the mei-yu front from moving south-

    ward and thereby prolongs the frontal rainfall along the

    lower reaches of the Yangtze River and the Huaihe River

    valleys (Chang et al. 2000a,b; Zhou and Yu 2005; Sui et al.

    2007; Wu and Zhou 2008).

    As a response to El Nino forcing, the WNPAC may

    play a role in the phase reversal of El Nino (Wang et al.

    2001). Easterly anomalies to the south of the WNPAC

    may stimulate oceanic upwelling Kelvin waves and thus

    reverse warm SST anomalies (SSTAs) in the equatorial

    central-eastern Pacific (Weisberg and Wang 1997a,b; Wang

    et al. 1999a; Kim and Lau 2001; Li et al. 2007; Ohba and

    Ueda 2009).

    The WNPAC is tightly coupled with underlying ocean

    surface cooling. The northeasterly anomalies over the

    southeastern flank of the WNPAC increase the back-

    ground mean northeasterly trades and generate a colder

    SST in situ through enhanced evaporation. The nega-

    tive SSTAs further suppress convection and stimulate a

    descending Rossby wave to the northwest and thus re-

    inforce the WNPAC (Wang et al. 2003). This positive

    windevaporationSST feedback maintains the WNPAC

    through the El Nino mature winter and the subsequent

    spring (Wang et al. 2000). The windevaporationSST

    feedback was first proposed to study the equatorial asym-

    metry of the intertropical convergence zone (Xie and

    Philander 1994; Li 1997), and then was used to study the

    decadal variability over the tropical Atlantic (Chang et al.

    1997; Xie 1999) and the phase locking of ENSO (Wang

    et al. 1999b). In the study, we will explore the nonlinear

    feature of the windevaporation process over the western

    North Pacific.

    While the WNPAC is maintained by local airsea

    interaction, it is initiated possibly through the following

    three routes. First, circulation anomalies in response to

    the El Nino heating over the equatorial central Pacific

    generate cold SSTAs in the western Pacific, which fur-

    ther set up the WNPAC (Wang et al. 2000). Second, the

    deepening of the East Asian trough and the intrusion

    of midlatitude cold air into the Philippine Sea might

    trigger the WNPAC (Wang and Zhang 2002; Lau and

    Nath 2006). Third, the WNPAC results from the east-

    ward movement of an anomalous anticyclone established

    over the northern Indian Ocean (Chou 2004; Chen et al.

    2007).

    Most previous studies assumed a symmetric circula-

    tion feature between El Nino and La Nina; namely,

    there is an anomalous anticyclone (cyclone) over the

    WNP during the El Nino (La Nina) mature winter. How-

    ever, it is important to note that El Nino and La Nina

    have a significant asymmetry in amplitude, structure, and

    temporal evolution (e.g., Hoerling et al. 1997; Burgers

    and Stephenson 1999; Kang and Kug 2002; Jin et al.

    2003; An and Jin 2004; An et al. 2005). The anomalous

    convection over the equatorial central Pacific during La

    Nina tends to shift to the west of its El Nino counterpart

    (Hoerling et al. 1997). Numerical model experiments

    indicated that this asymmetry is attributed to nonlinear

    atmospheric responses to the underlying SSTA (Hoerling

    et al. 1997; Kang and Kug 2002).

    The anomalous convective heating over the equatorial

    central Pacific is a crucial factor that impacts the circu-

    lation over the WNP (Lau and Nath 2000; Wu and Zhou

    2008; Zhou et al. 2009a,b). Given the asymmetric SSTA

    pattern between El Nino and La Nina, one may wonder

    whether the WNP atmospheric response to El Nino and

    La Nina is asymmetric. In this paper, we attempt to ad-

    dress the following questions: 1) are WNPAC and WNP

    cyclone (WNPC) during the El Nino and La Nina ma-

    ture winter asymmetric and 2), If they show an asym-

    metric characteristic, what are the physical mechanisms

    that cause the asymmetry?

    The rest of the paper is organized as follows. Datasets,

    analysis methods, and an atmospheric general circula-

    tion model (AGCM) are described in section 2. Section

    3 presents the asymmetric circulation features over the

    WNP between El Nino and La Nina. In section 4, we

    discuss two possible factors that cause the asymmetry.

    The two factors are further examined through a series of

    AGCM numerical experiments in section 5. The impacts

    of the asymmetry on ENSO evolution are discussed in

    section 6. Summary and concluding remarks are given in

    section 7.

    2. Data, method, and model experiments

    a. Data and method

    The datasets used in the present study consist of

    1) the 850-hPa and 1000-hPa wind fields and the surface

    heat flux from the National Centers for Environment

    PredictionNational Center for Atmospheric Research

    (NCEPNCAR) reanalysis data (Kalnay et al. 1996)

    for the period from 1948 to 2004, 2) SST data from the

    Hadley Centre Global Sea Ice and Sea Surface Tem-

    perature dataset (HadISST) (Rayner et al. 2003) for

    the period 19482004, 3) monthly precipitation anoma-

    lies over global land and oceans for the same period

    from the precipitation reconstruction (PREC) dataset

    (Chen et al. 2002), and 4) oceanic subsurface temperature

    data from a simple ocean data assimilation analysis of

    global upper ocean (SODA) (Carton et al. 2000) for the

    period from 1950 to 2000.

    4808 J O U R N A L O F C L I M A T E VOLUME 23

  • A composite analysis method is applied. We chose nine

    El Nino events and nine La Nina events in the period

    19482002. The selection of El Nino and La Nina events

    is based on the threshold of one standard deviation of

    wintertime [DecemberFebruary (DJF)] mean Nino-3.4

    index, which is defined as the area-averaged SSTA over

    the region 58N58S, 12081708W. The selected ENSOyears are listed in Table 1.

    To focus on the interannual time scale, variations

    longer than 8 years are filtered out from the original

    datasets with a Lanczos filter (Duchon 1979). Following

    Hoerling et al. (1997), the difference between the com-

    posite El Nino and La Nina events is regarded as a sym-

    metric component, and the sum of them is regarded as

    an asymmetric component.

    b. Model description and experiment design

    The AGCM used in the study is the ECHAM version

    4.6 (ECHAM4) developed by the Max Planck Institute

    for Meteorology (Roeckner et al. 1996). The model was

    run at a horizontal resolution of spectral triangular 42

    (T42), roughly equivalent to 2.88 latitude 3 2.88 longi-tude, with 19 vertical levels in a hybrid sigma pressure-

    coordinate system extending from the surface to 10 hPa.

    The SST data used as the lower boundary condition is

    the same as that used in the observational analysis.

    As listed in Table 2, four sets of model experiments

    are designed:

    1) Control run (CTRL run):the ECAHM4 was inte-

    grated for 10 years, forced by monthly climatologi-

    cal SST based on the period 19482005.

    2) Global SSTA forcing runs for El Nino and La Nina:

    composite El Nino or La Nina SSTA was added to

    the climatological SST in the global ocean (GB ex-

    periment); the run with SSTA derived from El Nino

    (La Nina) is referred to as the GBEL (GBLA) run.

    3) Tropical central-eastern Pacific SSTA forcing runs

    for El Nino and La Nina: These experiments were the

    same as the GB experiments except that only the

    positive (negative) SSTA in the tropical central-eastern

    Pacific (CE experiment) associated with El Nino (La

    Nina) was added to the climatological mean SST

    field. The runs with SSTA derived from El Nino (La

    Nina) are referred to as CEEL (CELA) runs.

    4) Western North Pacific SSTA forcing runs for El Nino

    and La Nina: the experiments were the same as the

    GB experiments except that only the negative (pos-

    itive) SSTA in the western North Pacific (WP exper-

    iment) associated with El Nino (La Nina) was added

    to the climatological mean SST field. The runs with

    SSTA derived from El Nino (La Nina) are referred to

    as WPEL (WPLA) runs.

    For each sets of experiment, an ensemble simulation

    with 10 members was performed. Among the 10 mem-

    bers, each realization only differed in initial condition

    but was forced by an identical SST field. The model was

    integrated from October to February in each simulation.

    The ensemble mean of DecemberFebruary is analyzed.

    3. Asymmetry of circulation anomalies betweenEl Nino and La Nina

    The composite 850-hPa streamfunction anomalies dur-

    ing the mature phase (DJF) of El Nino and La Nina

    are shown in Figs. 1a,b. In the El Nino composite, an off-

    equatorial anomalous anticyclone is evident over the

    WNP, with a center located in the Philippine Sea. The

    anomalous easterlies in the southern flank of the anti-

    cyclone extend eastward to 1558E. Twin cyclone coupletsstraddle over the central equatorial Pacific, accompanied

    with westerly anomalies on the equator. In contrast, in

    the La Nina composite the northern branch of the twin

    anticyclone couplets over the central equatorial Pacific

    TABLE 1. El Nino and La Nina events in the period of 19492002,

    used in the study.

    El Nino 1957, 1965, 1972, 1982, 1986, 1991, 1994, 1997, 2002

    La Nina 1949, 1955, 1970, 1973, 1975, 1984, 1988, 1998, 1999

    TABLE 2. List of numerical experiments.

    Expt SST forcing field Integration

    CTRL Global climatological SST 10 years

    GBEL and GBLA Add the composite SSTAs to the climatological

    SST of the global ocean

    10 realizations for both GBEL and GBLA

    CEEL and CELA Add the composite positive (negative) SSTAs in the tropical

    central-eastern Pacific (1608E1008W, 308S308N) to theclimatological SST in CEEL (CELA)

    10 realizations for both CEEL and CELA

    WPEL and WPLA Add the composite negative (positive) SSTAs in the

    WNP (12081608E, 08208N) to the climatological SST inWPEL (WPLA)

    10 realizations for both WPEL and WPLA

    15 SEPTEMBER 2010 W U E T A L . 4809

  • shift westward by 158 longitude with pronounced equa-torial easterly anomalies extending to 1358E. As a re-sult, the WNPC shifts westward relative to the WNPAC

    during El Nino, with a center located in the South China

    Sea (SCS).

    The asymmetric component of the low-level circula-

    tions (Fig. 1c) is characterized by an anomalous anticy-

    clone over the WNP, cyclone couplets over the equatorial

    central-eastern Pacific, and strong southwesterly winds

    from the Bay of Bengal to southeastern China. Note that

    the asymmetry between WNPAC and WNPC is the most

    significant feature in the asymmetric component of the

    low-level anomalous circulation fields between El Nino

    and La Nina. In contrast, the symmetric component

    (Fig. 1d) represents the averaged condition of El Nino

    and La Nina, with an anticyclone located between the

    composite WNPAC and the composite WNPC.

    As the maximum asymmetry of the 850-hPa stream-

    function appears over the WNP (28208N, 12581558E;the rectangular box in Fig. 1c), we calculate the box-

    averaged vorticity for each event, shown as a scatter

    diagram in Fig. 2, together with corresponding Nino-3.4

    indices. Nearly all El Nino events show negative vorticity

    anomalies over the WNP, except for the 1986 event. If the

    circulation anomalies were symmetric between El Nino

    and La Nina, we would expect positive vorticitiy anom-

    alies during La Nina. However, only four out of nine

    La Nina events show positive vorticity anomalies. This

    asymmetry is consistent with the fact that the anomalous

    cyclones during most La Nina events shift westward away

    from the WNP. Figure 2 shows that there is no significant

    linear relationship between the amplitude of Nino-3.4 in-

    dices and the WNP vorticity. We explore the cause of the

    circulation asymmetry in next section.

    4. Mechanisms responsible for the asymmetrybetween WNPAC and WNPC

    The anomalous circulation over the WNP during the

    ENSO mature winter results from both local forcing of

    a negative SSTA in the WNP and remote forcing of a

    positive SSTA in the equatorial central-eastern Pacific

    FIG. 1. Composite DJF mean 850-hPa streamfunction anomalies

    for (a) nine El Nino events and (b) nine La Nina events, listed in

    Table 1 (units: 106 m2 s21). (c) Asymmetric component estimated

    by the sum of (a) and (b). The shading represents the 10% signif-

    icance level. (d) Symmetric component estimated by the difference

    of (a) and (b). Positive (solid lines) and negative values (dashed

    lines) are denoted: contour interval 0.5.

    FIG. 2. Scatter diagram of the DJF mean Nino-3.4 index (8C)vs the box-averaged vorticity anomalies over the western North

    Pacific (28208N, 12581558E, rectangular box in Fig. 1, units:1026 s21) for 18 ENSO events listed in Table 1: El Nino (dots) and

    La Nina (crosses) events.

    4810 J O U R N A L O F C L I M A T E VOLUME 23

  • (Wang et al. 2000; Lau and Nath 2000, 2003; Lau et al.

    2004). To explore the asymmetric characteristics of the

    remote and local forcing between El Nino and La Nina,

    we show composite precipitation and SST anomalies and

    their asymmetric components in Fig. 3. Positive precipi-

    tation anomalies associated with El Nino extend from

    the equatorial central Pacific to the far eastern Pacific.

    In contrast, negative precipitation anomalies associated

    with La Nina are restricted to the west of 1408W andtheir center is located farther west, compared with the

    El Nino counterpart. Their asymmetric component ex-

    hibits a zonal dipole pattern, with a positive (negative)

    pole over the equatorial central-eastern (western) Pacific

    (Fig. 3f).

    The SSTAs over the tropical central-eastern Pacific

    also present an asymmetric feature. The positive SSTAs

    in the tropical eastern Pacific during El Nino are stron-

    ger than the negative SSTAs during La Nina, whereas

    the negative SSTAs during La Nina are stronger in the

    tropical central Pacific and extend farther westward. As

    a result, the asymmetric component of the SSTAs also

    shows a zonal dipole pattern in the equatorial Pacific

    (Fig. 3e).

    The above results imply that the asymmetry of the

    WNP circulation anomalies may be partially caused by

    the zonal asymmetries of the precipitation and SST

    anomalies between El Nino and La Nina. The anoma-

    lous SST and precipitation centers shift farther west-

    ward during La Nina. It is likely that the westward shift

    of the precipitation anomalies during La Nina causes

    the twin anomalous anticyclone couplets expanding into

    the tropical western Pacific. As a result, the center of the

    WNPC is pushed farther to the SCS (Fig. 1b).

    To illustrate the effect of the zonal asymmetry of the

    anomalous precipitation center between El Nino and

    La Nina on the WNP circulations, a scatter diagram of

    the WNP vorticity versus the longitudinal location of the

    anomalous precipitation center over the equatorial cen-

    tral Pacific is shown in Fig. 4. Five out of nine El Nino

    events have positive convection centers located east of

    1708W, while nearly all La Nina events have negativeconvection centers located near the date line. It is in-

    teresting to note that, when the precipitation centers are

    located to the east of 1708W, there is a close relationshipbetween the sign of the WNP vorticity anomalies and the

    SSTA in the equatorial central-eastern Pacific; namely, an

    FIG. 3. As in Figs. 1ac but for (a),(c),(e) SST anomalies (8C) and (b),(d),(f) precipitation anomalies (mm day21).

    15 SEPTEMBER 2010 W U E T A L . 4811

  • anticyclone (cyclone) is associated with El Nino (La Nina).

    However, when the anomalous precipitation centers

    shift to the west of 1708W, the relationship becomescomplicated, suggesting that some other factors rather

    than the longitudinal location of the heating must play

    active roles.

    In addition to the equatorial eastwest asymmetry, a

    significant asymmetry of SSTAs also appears in the off-

    equatorial WNP (Fig. 3e). The cold SSTAs in the WNP

    during El Nino are much stronger than the warm SSTAs

    during La Nina. Due to the weaker SSTAs, the pre-

    cipitation anomalies over the WNP during La Nina are

    also much weaker than that during El Nino (Figs. 3b,d).

    There are two negative precipitation centers during

    El Nino, located at 158N, 1208E and 58N, 1508E, respec-tively (Fig. 3a). The former has a mirror image during

    La Nina, but the latter does not (Fig. 3b). It is crucial to

    understand why the asymmetric SSTAs, which cause

    asymmetric precipitation anomalies, emerge in the WNP.

    To clearly illustrate the asymmetric evolution of the

    SSTAs in the WNP (Fig. 3e), we define a western North

    Pacific SST (WNPSST) index as the area-averaged SSTAs

    in the region 28158N, 13081558E. Figure 5 shows theevolution of the WNPSST index during El Nino and

    La Nina. The WNPSST index reaches a peak phase in

    January, slightly lagging the mature phase of ENSO

    (Rasmusson and Carpenter 1982). It experiences two

    fast growing stages, one in JuneAugust (JJA) and the

    other in NovemberDecember (NDJ). During both stages,

    the absolute values of WNPSST index tendencies during

    El Nino are much larger than that during La Nina. Thus,

    the amplitude of the WNPSST in DJF during El Nino is

    about twice as large as that during La Nina.

    Why does the WNPSST experience a stronger (weaker)

    growth during the El Nino (La Nina) development sum-

    mer? To explore the possible mechanisms for the asym-

    metric evolutions of the WNPSST indices in JJA, we

    show composite SST, 1000-hPa wind, and precipitation

    anomalies during El Nino and La Nina in Fig. 6. Note

    that the warm SSTAs already have been established in

    the equatorial central-eastern Pacific during the El Nino

    development summer (Fig. 6b). The warm SSTAs en-

    hance convection over the equatorial central Pacific

    (Fig. 6a), and thus stimulate twin low-level cyclone cou-

    plets in both sides of the equator, consistent with the

    classical MatsunoGill pattern (Matsuno 1966; Gill 1980).

    The northern branch of the cyclonic circulation anoma-

    lies is stronger than the southern one, which can be in-

    ferred from northward cross-equatorial flow anomalies

    over the Maritime Continent. The asymmetry is possibly

    attributed to the hemispheric asymmetry of the boreal

    summer mean flow such as the easterly vertical shear

    (Wang et al. 2003). The asymmetric flow over the Mari-

    time Continent may be further enhanced by airsea in-

    teraction in the tropical southeastern Indian Ocean (Li

    et al. 2002; Li et al. 2003; Wu et al. 2009).

    The pattern of SSTAs during La Nina is generally

    symmetric with respect to that during El Nino (Fig. 6d).

    FIG. 4. Scatter diagram of the longitudinal location of the max-

    imum amplitude of DJF mean 108N108S averaged precipitationanomalies over the Pacific vs the box-averaged vorticity anomalies

    over the western North Pacific (28208N, 12581558E; units: 1026 s21)for 18 ENSO events, listed in Table 1: El Nino (dots) and La Nina

    (crosses) events.

    FIG. 5. Monthly mean time series (three-point smoother) of the

    WNPSST index, (solid line, 8C) and the temporal evolution ofWNPSST tendency (dashed line, 8C month21) for the (a) El Ninoand (b) La Nina composites. The WNPSST index is defined as box-

    averaged SST anomalies in the region 28158N, 13081558E; Y(0)and Y(1) represent ENSO-developing and ENSO-decaying years,

    respectively.

    4812 J O U R N A L O F C L I M A T E VOLUME 23

  • In particular, in the WNP, the amplitudes of the cold

    SSTAs during El Nino and the warm SSTAs during

    La Nina are quite close. Corresponding to the generally

    symmetric SSTA patterns, the precipitation and surface

    wind anomalies in the WNP between El Nino and

    La Nina are also quite symmetric (Figs. 6a,c).

    Given the symmetric SST and wind patterns between

    El Nino and La Nina, a natural question that needs be

    addressed is what causes the asymmetric SSTA tendency

    in JJA. In the following we argue that the SSTA tendency

    asymmetry is primarily attributed to the asymmetry of the

    surface latent heat flux (LHF) anomaly. The LHF may be

    expressed as

    LHF 5 r0L

    yC

    eU(Q

    sQ

    a), (1)

    where ro, Ly, Ce, and U denote air density, the latent

    heat of vaporization, the surface exchange coefficient, and

    surface wind speed, respectively; Qs is saturated specific

    humidity at SST and Qa is air specific humidity at 10 m.

    The anomalous surface wind speed and LHF fields

    during the El Nino and La Ninadeveloping summers

    and their asymmetric components are shown in Fig. 7.

    The similarity of the patterns for anomalous wind and

    LHF suggests that the surface wind speed anomaly dom-

    inates the LHF anomaly in the WNP. Note that there is

    a large asymmetric component in the anomalous LHF

    field between El Nino and La Nina (Fig. 7f). The asym-

    metric LHF component is positive over the key region of

    the WNP (28158N, 13081558E), implying a stronger

    cooling (weaker warming) tendency in WNP during

    El Nino (La Nina).

    Why are the anomalous surface wind speed and LHF

    fields greatly asymmetric in WNP, while the surface wind

    anomalies are quite symmetric in JJA between El Nino

    and La Nina (Figs. 6a,c)? Below we construct a simple

    ideal theoretical model to understand the nonlinear rela-

    tionship between the LHF anomaly and the wind anomaly

    (Wang et al. 1999b).

    Consider an idealized case in which only the mean and

    anomalous zonal winds are involved and the mean wind

    is westerly. Denote u and u9 as the magnitude of the meanand anomalous winds, and sps (spo) as the wind speed

    anomaly when the anomalous zonal wind is in the same

    (opposite) direction with the mean wind. Then, we have

    sps 5 u 1 u9j j u (2)

    and

    spo 5 u u9j j u. (3)

    In the case when the amplitude of the mean wind is

    stronger than that of the anomalous wind (i.e., u . u9),one may obtain a symmetric anomalous windwind speed

    relationship:

    sps 5 u9, (4)

    spo 5u9. (5)

    FIG. 6. Composite JJA mean 850-hPa wind anomalies (vector, units: m s21) and precipitation anomalies (shading,

    mm day21) for (a) El Nino and (c) La Nina events. Composite JJA mean SSTAs (8C) for (b) El Nino and (d) La Ninaevents.

    15 SEPTEMBER 2010 W U E T A L . 4813

  • In the case when the amplitude of the mean wind is

    weaker than that of the anomalous wind (i.e., u , u9),one may obtain an asymmetric wind speed relationship as

    sps 5 u9, (6)

    and

    spo 5 u9 2u. (7)

    Considering an extreme condition for which the mean

    wind is zero, the total wind speed would increase, no

    matter what direction the anomalous wind blows.

    The theoretic model above indicates that there exist

    two dynamic regimes in the evaporationwind feedback.

    The key difference between the two regimes lies in the

    relative amplitudes of the mean and anomalous winds.

    In the key region of the WNP (28158N, 13081558E), the

    amplitude of the anomalous zonal wind is about 3 m s21.

    However, the strength of the mean zonal wind is much

    smaller than 3 m s21 in most of the region, except in the

    northeastern corner of the region (Fig. 8a). Area-averaged

    zonal wind in the region is about 0.7 m s21.

    Figure 8b shows the theoretical model result based on

    the parameter values derived from the observation (i.e.,

    the mean wind of 0.7 m s21 and the anomalous wind

    amplitude of 3 m s21). When the anomalous wind is in

    the range from 20.7 to 0.7 m s21 (i.e., weaker than themean wind), resultant wind speed anomalies are sym-

    metric. A westerly anomaly leads to a positive wind speed

    anomaly, whereas an easterly anomaly leads to a nega-

    tive wind speed anomaly (Fig. 8b). When the anomalous

    wind is beyond the range of 20.7 to 0.7 m s21, resultantwind speed anomalies are asymmetric. Both strong west-

    erly and easterly anomalies lead to a positive wind speed

    (and thus LHF) anomaly (Fig. 8b).

    FIG. 7. Composite JJA mean 1000-hPa wind speed anomalies (m s21) for (a) El Nino and (c) La Nina events and

    (b),(d) corresponding latent heat flux anomalies (W m22). (e) Asymmetric component of the wind speed anomalies

    estimated by the sum of (a) and (c). (f) Asymmetric component of the latent heat flux anomalies estimated by the sum

    of (b) and (d).

    4814 J O U R N A L O F C L I M A T E VOLUME 23

  • To compare the theoretical solution with real data, we

    show a scatter diagram of the zonal wind anomaly versus

    the wind speed anomaly for each grid in the WNP (28158N, 13081558E; see Fig. 8b). As shown in Fig. 6, theanomalous winds in the region are dominated by their

    zonal component. Thus, we use the zonal wind anoma-

    lies to represent the total anomalous winds. The rela-

    tionship between the zonal wind anomalies and wind

    speed anomalies is quite consistent with the theoretical

    model; that is, the enhancement of both positive and

    negative zonal wind anomalies during El Nino and

    La Nina leads to the increase of the wind speed and LHF

    anomalies. The positive LHF anomaly enhances the

    WNP SST cooling during El Nino but weakens the WNP

    SST warming during La Nina. Thus, the nonlinear re-

    lationship causes the asymmetry of the SST tendency in

    the WNP between El Nino and La Nina.

    The analyses above indicate that the area 28158N,13081558E is a key region for generating asymmetricLHF anomalies between El Nino and La Nina during

    their development summer. It is located in a transitional

    zone between the monsoon westerly and the easterly

    trade wind, where the mean zonal wind is very weak

    (Fig. 8a). Even though the easterly anomalies during

    La Nina and the westerly anomalies during El Nino in JJA

    are approximately symmetric (Figs. 6a,c), the wind speed

    anomalies are enhanced in both cases. The asymmetry

    of LHF anomalies caused by the wind speed asymmetry

    may further induce the asymmetric SSTA tendencies in

    the WNP between El Nino and La Nina.

    Another WNPSST tendency asymmetry between

    El Nino and La Nina occurs in NDJ (Fig. 5), which is

    primarily attributed to the asymmetry of anomalous

    wind fields. Note that in the WNP, the easterly anomaly

    appears in both El Nino and La Nina composites (Figs.

    9a,b). The northeasterly anomaly over the WNP during

    El Nino connects the WNPAC to the west and a cyclonic

    circulation anomaly to the east (Fig. 9a). The former is

    a Rossby wave response to the local negative heating

    anomaly, whereas the latter is likely a result of a Rossby

    wave response to the positive heating anomaly over the

    equatorial central Pacific. The easterly anomaly over the

    WNP during the La Nina, on the other hand, is the part

    of the anomalous circulation of the twin anticyclone

    couplet that is a direct response to the negative heating

    over the central equatorial Pacific (Fig. 9b). As the mean

    wind is northeasterly in the region, the anomalous wind

    would lead to an enhanced wind speed and LHF during

    both El Nino and La Nina (figure not shown). Corre-

    spondingly, the asymmetric components of the surface

    wind speed and the LHF anomalies are positive in the

    WNP region (28158N, 13081558E; see Fig. 9c). Theasymmetry of LHF anomalies leads to an enhanced

    cooling during El Nino but a reduced warming during

    La Nina in this region. As a result, the asymmetry of the

    WNP SSTAs between El Nino and La Nina becomes

    even greater, leading to a more significant difference

    between the WNPAC and the WNPC.

    5. The results of AGCM experiments

    Observational analyses above suggest that both asym-

    metries of the remote forcing from the tropical central

    Pacific and the local SST forcing in the WNP may cause

    the asymmetry between WNPAC and WNPC. In this

    section we examine their relative roles by analyzing the

    results of the numerical experiments listed in Table 2.

    The GB experiment is analyzed first to test the basic

    model skill in reproducing the asymmetry between

    WNPAC and WNPC during the ENSO mature winter

    FIG. 8. (a) Climatological JJA mean 1000-hPa wind. The box

    represents the target region of the WNP (28158N, 13081558E). (b)Scatter diagram of the JJA mean 1000-hPa zonal wind anomalies vs

    the 1000-hPa wind speed anomalies for each grid in the target re-

    gion of WNP in 18 ENSO events. The thick line is calculated based

    on a theoretical model in which the mean wind is set to 0.7 m s21.

    15 SEPTEMBER 2010 W U E T A L . 4815

  • (DJF). The simulated precipitation and 850-hPa wind

    anomalies in the GBEL and GBLA runs and their asym-

    metric components are shown in Fig. 10. The anomalies

    are calculated as departures from the CTRL run. The

    simulated circulation and precipitation anomalies can be

    compared against Fig. 1 and the right panel of Fig. 3,

    respectively. The GB experiment reproduces the WNPAC

    (WNPC) during El Nino (La Nina) and corresponding

    negative (positive) precipitation anomalies over the WNP

    and positive (negative) precipitation anomalies over the

    equatorial central Pacific. The model reproduces the

    major asymmetric characteristic between the WNPAC

    and WNPC; that is, the WNPC is much weaker than

    the WNPAC and is located farther to the west of the

    WNPAC. It should be noted that the simulated asym-

    metric component between WNPAC and WNPC is much

    larger that that in the observation, owing to the weaker

    WNPC in the GBLA runs. The model reproduces well

    the zonal shift of the anomalous heating over the equa-

    torial central Pacific and the strength difference of the

    anomalous heating over the WNP, indicating that it is

    capable of capturing the impacts of both remote central-

    eastern Pacific and local WNP SST forcing. The results

    give us confidence to further explore their relative roles

    through sensitive experiments.

    The precipitation and 850-hPa wind anomalies simu-

    lated by the CEEL and CELA runs and their asym-

    metric components are shown in Figs. 11ac. The CEEL

    FIG. 9. (a) El Nino and (b) La Nina composite NDJ mean

    1000-hPa wind anomalies (vectors, m s21) and precipitation

    anomalies (contours, mm day21). (c) Asymmetric components of

    1000-hPa wind speed anomalies (contours, m s21) and latent heat

    flux anomalies (shading, W m22) between El Nino and La Nina.

    FIG. 10. (a) Difference of the precipitation (shading, mm day21)

    and 850-hPa wind (vectors, m s21) fields in DJF between the

    GBEL and CTRL runs. (b) Same as (a), but for the GBLA runs. (c)

    Asymmetric component of (a) and (b), estimated by their sum; only

    values in the 10% significance level are shown.

    4816 J O U R N A L O F C L I M A T E VOLUME 23

  • (CELA) reproduces the WNPAC (WNPC) and the cor-

    responding negative (positive) precipitation anomalies

    over the WNP and positive (negative) precipitation

    anomalies over the equatorial central Pacific, though

    some biases exist. Compared with the GBEL runs, the

    WNPAC simulated by the CEEL runs shifts westward

    to some extent, and the negative precipitation anomalies

    over the WNP are weaker, especially in the region of

    13081408E. The differences between CELA and GBLAare larger. The easterly anomalies induced by the negative

    heating over the equatorial central Pacific extend into the

    Maritime Continent. The WNPC shifts farther northward.

    The differences occur because of the lack of local forcing

    in the WNP. In summary, the asymmetry between the

    WNPAC and WNPC generally can be reproduced in the

    CE experiments, although the center of the asymmetric

    component shifts westward relative to the GB experiments.

    The results of the WP experiments are shown in Figs.

    11df. The WNPAC is reproduced, but with amplitudes

    far weaker than their counterparts in the CEEL runs. In

    the CELA runs, the WNPC is weakly simulated, with

    weak westerlies seen in the WNP. However, compared

    with the CE experiments, the locations of the WNPAC

    in the CEEL runs and the WNP westerly anomalies in

    the CELA runs are closer to those in the GB experi-

    ments. The asymmetry of the WNP circulation anoma-

    lies is also reproduced by the WP experiments, though

    the asymmetric component is weaker than that in the

    GB and CE experiments.

    The model results support our hypothesis based on

    observational diagnosis that both of the asymmetries of

    the remote forcing from the equatorial central-eastern

    Pacific and the local SSTA forcing in the WNP contribute

    to the asymmetry between the WNPAC and WNPC. The

    FIG. 11. As in Fig. 10 but for the (a)(c) CEEL and CELA runs and (d)(f) WPEL and WPLA runs.

    15 SEPTEMBER 2010 W U E T A L . 4817

  • numerical experiments demonstrate that the remote forc-

    ing from the equatorial central-eastern Pacific play greater

    roles than the local WNP SST.

    6. Impact of the asymmetry on ENSO phasereversal

    Previous studies noted the essential role of equatorial

    zonal wind stress over the equatorial western Pacific in

    ENSO phase reversal (Weisberg and Wang 1997a,b; Kim

    and Lau 2001; Lau and Wu 2001; Kug and Kang 2006).

    To investigate relationship between the WNPC/WNPAC

    and the zonal wind stress anomalies over the equatorial

    western Pacific, we show the scatter diagram of the box-

    averaged zonal pseudowind stress anomalies over the

    equatorial western Pacific (58S58N, 12581558E) versus850-hPa vorticity anomalies over the WNP (28208N,12581558E; see Fig. 12). Their correlation coefficientreaches 0.79, which is statistically significant at the 1%

    level, indicating that the WNP vorticity accounts for

    about half of the variance of the zonal wind stress over

    the equatorial western Pacific during the ENSO mature

    winter.

    The easterly wind stress anomalies prevail during

    seven out of nine El Nino events, except for the 1986 and

    1994 events. In contrast, only three La Nina events have

    westerly wind stress anomalies at the equator. Some

    La Nina events even have strong equatorial easterly

    wind stress anomalies.

    The asymmetry of the equatorial wind stress anoma-

    lies over the western Pacific is likely responsible for

    the asymmetry of ENSO evolution (Ohba and Ueda

    2009). Figure 13 shows that following the El Nino and

    La Nina peak phase in boreal winter, canonical El Nino

    experiences a fast decay and translates to a negative

    phase in boreal summer. In contrast, La Nina tends to

    persist through the next year. To quantify the decaying

    rate of the Nino-3.4 index, we calculated the tendency

    of the Nino-3.4 indices of composite El Nino and

    La Nina, respectively (Fig. 13). Both El Nino and La Nina

    reach the fastest decaying rate in March, with the for-

    mer being much larger than the latter. In addition,

    El Nino tends to keep the strong decay rate much

    longer than La Nina. As an exception, the 1986 El Nino

    does not decay after DJF and is maintained until the

    1988 summer.

    To show the different evolutional characteristics of

    oceanic subsurface during El Nino and La Nina under

    asymmetric forcing of the equatorial zonal wind stress

    anomalies over the western Pacific, we show the temporal

    evolution of composite 208C isotherm depth anomaliesalong the equator (Fig. 14). The change of the subsurface

    temperature is primarily caused by the thermocline var-

    iation, which is intimately linked to the passage of Kelvin

    waves. In September, two months before the ENSO

    mature phase, the 208C isotherms in the western Pacificduring El Nino and La Nina are both close to the cli-

    matological state. In December, after the formation of

    the WNPAC and the associated equatorial easterly wind

    stress anomalies over the western Pacific, the underlying

    thermocline during El Nino significantly rises (Fig. 14a),

    and the subsurface temperature in the equatorial west-

    ern Pacific decreases (figure not shown). In contrast, the

    amplitude of the thermocline depth anomalies during

    La Nina changes much less (Fig. 14b). In the course of

    eastward propagation, the upwelling Kelvin waves dur-

    ing El Nino are persistently stronger than the down-

    welling Kelvin waves during La Nina. Thus, El Nino

    decays much more quickly than La Nina.

    7. Summary and concluding remarks

    This paper deals with the asymmetry of the WNP at-

    mospheric circulation anomalies between the El Nino

    and La Nina mature winter. The physical mechanisms

    responsible for the asymmetry are studied from two

    plausible aspects: the intrinsic asymmetries of ENSO

    remote forcing and local airsea interaction. Both pro-

    cesses are demonstrated with idealized numerical ex-

    periments. The impacts of the asymmetry on the ENSO

    phase reversal are also investigated. The major findings

    are summarized below.

    FIG. 12. Scatter diagram of the DJF mean box-averaged zonal

    pseudowind stress anomalies over the equatorial western Pacific

    (58N58S, 12581558E) vs the box-averaged 850-hPa voriticityanomalies over the western North Pacific (28208N, 12581558E)for ENSO events, listed in Table 1: El Nino (dots) and La Nina

    (crosses) events.

    4818 J O U R N A L O F C L I M A T E VOLUME 23

  • d The WNPC during La Nina tends to shift westward

    relative to the WNPAC during El Nino. While the

    WNPAC center is located over the Philippine Sea, the

    WNPC center is located over the SCS. Correspond-

    ingly, equatorial westerly anomalies in the southern

    flank of the WNPC withdraw westward about 208 rel-ative to their El Nino counterparts.

    d Two possible mechanisms responsible for the phase

    shift are examined. First, as a Rossby wave response

    to the negative heating over the equatorial central

    Pacific, twin anticyclone couplets during La Nina shift

    westward relative to twin cyclone couplets during

    El Nino due to the spatial phase shift of SST and pre-

    cipitation anomalies over the equatorial central Pacific.

    Second, the intensities of negative SSTAs and associ-

    ated negative precipitation anomalies in the WNP dur-

    ing El Nino are stronger than the La Nina counterparts.

    The asymmetry of the SSTAs arises from asymmetric

    SSTA tendency in JJA and NDJ.d In JJA, though SSTAs, precipitation, and wind anom-

    alies are generally symmetric between El Nino and La

    Nina (i.e., a mirror image), surface wind speed anom-

    alies and corresponding latent heat flux are highly

    asymmetric. The asymmetric latent heat flux further

    causes the asymmetric SSTA tendency. In NDJ, the

    asymmetric SSTA tendency is primarily attributed to the

    asymmetry of anomalous wind fields between El Nino

    and La Nina; the WNP is covered by easterly anomalies

    during both El Nino and La Nina. The easterly anom-

    alies would enhance surface wind speed and latent heat

    flux. As a result, the asymmetry of the local SSTAs

    between El Nino and La Nina is further amplified.d Two distinctive windevaporation feedback regimes

    are identified to explain the asymmetry of the latent

    FIG. 13. Temporal evolution of monthly mean Nino-3.4 (58S58N, 19082408E) SST anom-alies (8C) from January of the ENSO development year to December of the following year for(a) El Nino and (b) La Nina events (thin lines). Composite Nino-3.4 SST anomalies (solid thick

    line) and the SSTAs tendency (dashed thick line) are also shown. Nino-3.4 indices are filtered

    by a five-point smother. Y(0) and Y(1) indicate the ENSO-development year and decay years,

    respectively.

    15 SEPTEMBER 2010 W U E T A L . 4819

  • heat flux anomaly during the ENSO development sum-

    mer (JJA). Since the wind speed anomalies are de-

    termined by both anomalous wind and background

    mean wind, when the mean wind is stronger (weaker)

    than the anomalous wind, wind speed anomalies are

    symmetric (asymmetric) with respect to the opposite

    anomalous winds. Therefore, the wind speed anoma-

    lies exhibit dominant asymmetric characteristics be-

    tween El Nino and La Nina in the transitional zone

    between the monsoon westerly and the trade easterly,

    where the mean wind is very weak. The asymmetry of

    the wind speed anomalies further cause the asymme-

    try of the latent heat flux anomalies.d The hypotheses based on data diagnosis are further

    confirmed by the results of numerical experiments,

    which indicate that both the asymmetries of the re-

    mote forcing from the equatorial central Pacific and

    the local SSTA forcing in the WNP contribute to the

    asymmetry between the WNPAC and WNPC. The

    impact of the former is larger than the latter in terms

    of the amplitudes of model responses.

    d Due to the westward shift of the WNPC center, westerly

    wind stress anomalies over the equatorial western Pa-

    cific during the La Nina mature phase are much weaker

    than easterly wind stress anomalies during El Nino. The

    weak (strong) wind stress anomalies during La Nina

    (El Nino) may stimulate weak (strong) oceanic Kelvin

    waves, leading to a slower (faster) phase transition to

    El Nino (La Nina). This may partially explain the

    evolution asymmetry between El Nino and La Nina.

    It should be noted that, in this study, we focus on the

    asymmetry of the WNP circulation anomalies between

    El Nino and La Nina, but pay less attention to the non-

    linearity between El Nino and La Nina events. For in-

    stance, the 1986 winter is an El Nino winter during which

    the WNP is covered by an anomalous cyclone, instead of

    an anomalous anticyclone. The SSTAs in the equatorial

    eastern Pacific did not decay after the winter, but per-

    sisted through the next year. The cause of this special

    event is not clear at present and warrants further studies.

    FIG. 14. Timelongitude distributions of 208C isotherm anomalies (m) averaged over 28N28S for the (a) El Ninocomposite and (b) La Nina composite [July(0)-June(1)]. The data are derived from the SODA reanalysis. Contour

    interval is 8; negative values are shaded.

    4820 J O U R N A L O F C L I M A T E VOLUME 23

  • While the endeavor in this regard has been puzzling

    owing to the limitation of sample size, a long-term coupled

    model simulation may help us to understand the issue.

    In addition, we only focus on the tropical processes in

    our analysis above. It was found that during the fall of

    El Nino events, the East Asian trough deepens, which

    may enhance cold air outbreaks and initiate the WNPAC

    (Wang and Zhang 2002). In fact, by checking the com-

    posite maps for the SeptemberOctober mean 500-hPa

    geopotential height anomalies for El Nino and La Nina,

    we note that the deepening of the East Asian trough

    during El Nino is much stronger than the shoaling of

    the trough during La Nina. This asymmetric character-

    istic may also contribute to the asymmetry between the

    WNPAC and WNPC. More detailed analysis of this

    midlatitude impact is needed in future work.

    Acknowledgments. This work was supported by NSFC

    Grants 40628006, 40821092, and 40523001, the National

    Basic Research Program of China (2006CB403603),

    RandD Special Fund for Public Welfare Industry (me-

    teorology; GYHY200706010, GYHY200806010). TL

    was supported by ONR Grants N000140710145 and

    N000140810256 and by the International Pacific Re-

    search Center, which is sponsored by the Japan Agency

    for Marine-Earth Science and Technology (JAMSTEC),

    NASA (NNX07AG53G), and NOAA (NA17RJ1230).

    REFERENCES

    An, S.-I., and F.-F. Jin, 2004: Nonlinearity and asymmetry of

    ENSO. J. Climate, 17, 23992412., Y. G. Ham, J. S. Kug, F. F. Jin, and I. S. Kang, 2005: El Nino

    La Nina asymmetry in the Coupled Model Intercomparison

    Project simulations. J. Climate, 18, 26172627.

    Burgers, G., and D. B. Stephenson, 1999: The normality of El Nino.

    Geophys. Res. Lett., 26, 10271030.

    Carton, J. A., G. Chepurin, and X. H. Cao, 2000: A Simple Ocean

    Data Assimilation analysis of the global upper ocean 195095.

    Part II: Results. J. Phys. Oceanogr., 30, 311326.Chang, C. P., Y. S. Zhang, and T. Li, 2000a: Interannual and in-

    terdecadal variations of the East Asian summer monsoon and

    tropical Pacific SSTs. Part I: Roles of the subtropical ridge.

    J. Climate, 13, 43104325.

    , , and , 2000b: Interannual and interdecadal varia-

    tions of the East Asian summer monsoon and tropical Pacific

    SSTs. Part II: Meridional structure of the monsoon. J. Climate,

    13, 43264340.

    Chang, P., J. Li, and H. Li, 1997: A decadal climate variation in

    the tropical Atlantic Ocean from thermodynamic airsea in-

    teractions. Nature, 385, 516518.Chen, J. M., T. Li, and C. F. Shih, 2007: Fall persistence barrier of

    sea surface temperature in the South China Sea associated

    with ENSO. J. Climate, 20, 158172.Chen, M., P. Xie, J. E. Janowiak, and P. A. Arkin, 2002: Global

    land precipitation: A 50-yr monthly analysis based on gauge

    observations. J. Hydrometeor., 3, 249266.

    Chou, C., 2004: Establishment of the low-level wind anomalies over

    the western North Pacific during ENSO development. J. Cli-

    mate, 17, 21952212.

    Duchon, C., 1979: Lanczos filtering in one and two dimensions.

    J. Appl. Meteor., 18, 10161022.

    Gill, A. E., 1980: Some simple solutions for heat-induced tropical

    circulation. Quart. J. Roy. Meteor. Soc., 106, 447462.

    Hoerling, M. P., A. Kumar, and M. Zhong, 1997: El Nino, La Nina,

    and the nonlinearity of their teleconnections. J. Climate, 10,

    17691786.

    Jin, F.-F., S.-I. An, A. Timmermann, and J. Zhao, 2003: Strong

    El Nino events and nonlinear dynamical heating. Geophys.

    Res. Lett., 30, 1120, doi:10.1029/2002GL016356.

    Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40-Year Re-

    analysis Project. Bull. Amer. Meteor. Soc., 77, 437471.

    Kang, I.-S., and J.-S. Kug, 2002: El Nino and La Nina sea surface

    temperature anomalies: Asymmetry characteristics associated

    with their wind stress anomalies. J. Geophys. Res., 107, 4372,

    doi:10.1029/2001JD000393.

    Kim, K. M., and K. M. Lau, 2001: Dynamics of monsoon-induced

    biennial variability in ENSO. Geophys. Res. Lett., 28, 315318.

    Kug, J.-S., and I.-S. Kang, 2006: Interactive feedback between

    ENSO and the Indian Ocean. J. Climate, 19, 17841801.Lau, K. M., and H. T. Wu, 2001: Principal modes of rainfallSST

    variability of the Asian summer monsoon: A reassessment of

    monsoonENSO relationship. J. Climate, 14, 28802895.

    Lau, N.-C., and M. J. Nath, 2000: Impact of ENSO on the vari-

    ability of the AsianAustralian monsoons as simulated in

    GCM experiments. J. Climate, 13, 42874309., and , 2003: Atmosphereocean variations in the Indo-

    Pacific sector during ENSO episodes. J. Climate, 16, 320.

    , and , 2006: ENSO modulation of the interannual and

    intraseasonal variability of the East Asian monsoonA

    model study. J. Climate, 19, 45084530., and H. Wang, 2004: Simulations by a GFDL GCM of ENSO-

    related variability of the coupled atmosphereocean system

    in the East Asian monsoon region. East Asian Monsoon,

    C.-P. Chang, Ed., World Scientific Series on Meteorology of

    East Asia, Vol. 2, World Scientific, 271300.

    Li, T., 1997: Airsea interactions of relevance to the ITCZ: The

    analysis of coupled instabilities and experiments in a hybrid

    coupled GCM. J. Atmos. Sci., 54, 134147.

    , Y. S. Zhang, E. Lu, and D. Wang, 2002: Relative role of

    dynamic and thermodynamic processes in the development

    of the Indian Ocean dipole. Geophys. Res. Lett., 29, 2110,

    doi:10.1029/2002GL015789.

    , B. Wang, C. P. Chang, and Y. S. Zhang, 2003: A theory for the

    Indian Ocean dipole-zonal mode. J. Atmos. Sci., 60, 2119

    2135.

    Li, Y., R. Lu, and B. Dong, 2007: The ENSOAsian monsoon in-

    teraction in a coupled oceanatmosphere GCM. J. Climate, 20,

    51645177.

    Matsuno, T., 1966: Quasi-geostrophic motions in the equatorial

    area. J. Meteor. Soc. Japan, 44, 2542.

    Ohba, M., and H. Ueda, 2009: Role of nonlinear atmospheric re-

    sponse to SST on the asymmetric transition process of ENSO.

    J. Climate, 22, 177192.

    Rasmusson, E. M., and T. H. Carpenter, 1982: Variations in tropical

    sea surface temperature and surface wind fields associated with

    the Southern Oscillation/El Nino. Mon. Wea. Rev., 110, 554584.

    Rayner, N. A., D. E. Parker, E. B. Horton, C. K. Folland,

    L. V. Alexander, D. P. Rowell, E. C. Kent, and A. Kaplan,

    2003: Global analyses of sea surface temperature, sea ice, and

    15 SEPTEMBER 2010 W U E T A L . 4821

  • night marine air temperature since the late nineteenth cen-

    tury. J. Geophys. Res., 108, 4407, doi:10.1029/2002JD002670.

    Roeckner, E., and Coauthors, 1996: The atmospheric general cir-

    culation model ECHAM-4: Model description and simulation

    of present-day climate. Max-Planck-Institute for Meteorology

    Rep. 218, 90 pp. [Available online at http://www.mpimet.mpg.de/

    fileadmin/publikationen/Reports/MPI-Report_218.pdf.]

    Sui, C.-H., P.-H. Chung, and T. Li, 2007: Interannual and inter-

    decadal variability of the summertime western North Pacific

    subtropical high. Geophys. Res. Lett., 34, L11701, doi:10.1029/

    2006GL029204.

    Wang, B., and Q. Zhang, 2002: Pacificeast Asian teleconnection.

    Part II: How the Philippine Sea anomalous anticyclone is es-

    tablished during El Nino development. J. Climate, 15, 3252

    3265.

    , R. G. Wu, and X. H. Fu, 2000: PacificEast Asian tele-

    connection: How does ENSO affect East Asian climate?

    J. Climate, 13, 15171536.

    , , R. Lukas, and S. I. An, 2001: A possible mechanism for

    ENSO turnabout. Dynamics of Atmospheric General Circu-

    lation and Climate, IAP/Academia Sinica, Ed., China Mete-

    orological Press, 552578.

    , , and T. Li, 2003: Atmospherewarm ocean interaction

    and its impacts on AsianAustralian monsoon variation.

    J. Climate, 16, 11951211.

    Wang, C. Z., R. H. Weisberg, and J. I. Virmani, 1999a: Western

    Pacific interannual variability associated with the El Nino

    Southern Oscillation. J. Geophys. Res., 104, 51315149.

    , , and H. J. Yang, 1999b: Effects of the wind speed

    evaporationSST feedback on the El NinoSouthern Oscil-

    lation. J. Atmos. Sci., 56, 13911403.

    Weisberg, R. H., and C. Z. Wang, 1997a: A western Pacific oscil-

    lator paradigm for the El Nino Southern Oscillation. Geophys.

    Res. Lett., 24, 779782.

    , and , 1997b: Slow variability in the equatorial west-

    central Pacific in relation to ENSO. J. Climate, 10, 19982017.

    Wu, B., and T. J. Zhou, 2008: Oceanic origin of the interannual and

    interdecadal variability of the summertime western Pacific

    subtropical high. Geophys. Res. Lett., 35, L13701, doi:10.1029/

    2008GL034584.

    , , and T. Li, 2009: Contrast of rainfallSST relationships

    in the western North Pacific between the ENSO developing

    and decaying summers. J. Climate, 16, 43984405.

    Wu, R., Z.-Z. Hu, and B. P. Kirtman, 2003: Evolution of ENSO-

    related rainfall anomalies in East Asia. J. Climate, 16, 3742

    3758.

    Xie, S.-P., 1999: A dynamic oceanatmosphere model of the

    tropical Atlantic decadal variability. J. Climate, 12, 6470.

    , and S. G. H. Philander, 1994: A coupled oceanatmosphere

    model of relevance to the ITCZ in the eastern Pacific. Tellus,

    46A, 340350.

    Zhang, R. H., and A. Sumi, 2002: Moisture circulation over East

    Asia during El Nino episode in northern winter, spring and

    autumn. J. Meteor. Soc. Japan, 80, 213227.

    , , and M. Kimoto, 1996: Impact of El Nino on the East

    Asian monsoon: A diagnostic study of the 86/87 and 91/92

    events. J. Meteor. Soc. Japan, 74, 4962.Zhou, T. J., and R. C. Yu, 2005: Atmospheric water vapor transport

    associated with typical anomalous summer rainfall pat-

    terns in China. J. Geophys. Res., 110, D08104, doi:10.1029/2004JD005413.

    , B. Wu, and B. Wang, 2009a: How well do atmospheric gen-

    eral circulation models capture the leading modes of the in-

    terannual variability of AsianAustralian monsoon? J. Climate,

    22, 11591173.

    , and Coauthors, 2009b: Why the western Pacific subtropical

    high has extended westward since the late 1970s. J. Climate, 22,

    21992215.

    4822 J O U R N A L O F C L I M A T E VOLUME 23

    Asymmetry_.pdf2010jcli3222-Wu-Li-ZHou.pdf